This proposal offers a unique opportunity to define the antibiotic treatment, patient characteristics and disease manifestations that affect otitis media outcome by using differences in national primary care practices. The natural history of acute otitis media (AOM) is incompletely known and the effectiveness of antibiotic therapy uncertain. The effects of patient characteristics, such as age, past history of AOM attacks, symptoms and tympanic membrane findings at the initial physician visit, and environmental variables, including day care attendance, on recovery are for the most part unknown. The international literature does not support the routine antibiotic treatment of AOM for 10 days that is standard in the United States. Nor does it identify patients likely to have poor outcomes using this treatment. The latter is a prerequisite to developing effective treatment strategies for those at high risk of poor outcomes, including recurrent otitis and chronic effusion. This quasi-experimental project will examine the separate and combined effects of antibiotic therapy, patient variables, day care, and signs and symptoms present at the initial physician visit for AOM on either the resolution of an attack or evidence of poor outcome at a follow-up exam in 3 primary care populations in the United States, the United Kingdom, and Netherlands. Forty-two primary care physicians from each country will recruit 13 patients age 3-12 months and 13 patients age 13-180 months who meet inclusion criteria. Data on the patient characteristics, including objective measurements made by portable tympanometry and pneumatic otoscopy will be collected at the initial visit and all subsequent visits, including a mandatory 2 months follow-up visit. The project will test 7 hypotheses: Patients age 3-12 months will have greater rates of poor outcome than those age 13-180 months; outcomes for patients who receive antibiotics will not differ from those who do not receive antibiotics; there will be no differences between groups who receive antibiotics for 5 or 10 days; patients with pustular ear discharge or a history or presence of tympanostomy tubes at the index visit will have higher rates of poor outcome than those in whom these variables are not present; rates of poor outcome will increase with the number of episodes of AOM or of serous otitis during the preceding 18 months; day care outside of the home increases the rates of poor outcome; and patients with middle ear effusion at the index visit will have greater rates of poor outcome than those in whom effusion is absent, when controlling for the several variables included in this study. A final hypothesis defines patient characteristics of those at low risk (<5%) of poor outcome (age 12 months, those who have had at most one episode of AOM which was at least 6 months ago, and who present without fever, pustular ear discharge, or a history or presence of tympanostomy tubes). A decision tree will be developed to identify the relative benefits in terms of health care utilization and costs, days of illness, and parental time lost from work associated with the 3 treatment strategies for patients with different risk characteristics. Otitis media is the most common pediatric morbidity leading to physician visits, and the 3rd most common cause for all U.S. physician visits. Annual costs of its treatment are over $1 billion; surgical care of complications costs an additional $1.2 billion annually in the U.S. The results of this study should improve our ability to individualize therapy for AOM patients and could result in significant cost savings by changing antibiotic prescribing practices.